1. Field of the Invention
The present invention relates generally to the field of measuring and monitoring the heating value of gaseous fuels such as natural gas and, more particularly, to a method and apparatus for determining the heating content or calorific value of such fuels from a determination substantially at the stoichiometric point of combustion of unknown mixtures containing precise volumetric proportions of fuel and air.
2. Description of the Prior Art
The heating value of gaseous fuels such as natural gas are frequently given assumed average numbers. Thus, the heating value of natural gas, for example, is frequently assumed to be 1000 British thermal units (BTU's) per cubic foot. In the past, pricing of such fuels has been based upon either the assumption of a nominal average value or the periodic checking of the actual value by a variety of time-consuming methods.
In one method chromatographic analysis of the constituents has been used to compute the actual heating value or BTU content of a given natural gas from the percentage composition of the mixture. In another method, the heat content has been determined by measuring the amount of heat liberated in burning exactly one cubic foot of gas (saturated with water) at standard conditions of temperature and pressure. The heat so liberated is absorbed by a weighed amount of water and the subsequent temperature rise of the water used to calculate the "gross or (higher) heating value." This "calorimeter bomb" approach, like the chromatographic approach, has several drawbacks.
Both methods involve reasonably expensive instrumentation and require considerable labor to perform the measurements and calculations. Such testing also, of necessity, introduces considerable time delay and certainly appears less desirable than an on-line system. Even so, these might be sufficient methods if, in fact, the composition of the gas being used did not vary greatly with time. However, the composition of natural gas, for example, may vary greatly in composition depending on the gas field from which it came and the treatment it receives before distribution. The gas that reaches the customer or consumer is frequently only about 85 percent methane with the remaining 15 percent being a mixture of various hydrocarbon molecules such as ethane, propane, n-butane, i-butane, etc. Also, as much as 25 percent of the gas reaching the customer may be made up of non-combustible constituents which occur naturally or have been added to the mixture. These include nitrogen air, and carbon dioxide. Natural gas is used herein as a representative example because it is by far the most widely used gaseous fuel.
The inerts, of course, add nothing to the heating value, and the heating value of alkane and other hydrocarbons of a higher order than methane have a higher heating value on a volumetric basis because of their higher molecular weights. In view of the great variation in constituents of natural gas, the heating value even in a single distribution system may vary greatly with time.
In addition to the other variations, numerous gas utilities have found that in severe winter weather it is cost justified to add a mixture of propane and air to the fuel in order to meet peak demand. They have found that the increased cost of propane at such peak load periods is less than the cost of additional distribution capacity which would otherwise be necessary to meet the peak loads with gas that cannot be stored as liquid adjacent to users.
As a result of all these factors, a random sampling of the heat content of the natural gas being distributed, might lead to great inaccuracies as to the actual heating value of the fuel delivered. Thus, it is necessary for the heat content of the natural gas to be continually monitored and adjusted in order to stay within promised specified limits and to assure that the user is charged for the proper amount of heating value he receives from the fuel.
Other incentives are involved in the desirability for providing an efficient on-line device for monitoring the heating value of gas. Users, especially those in industries, which require large amounts of gas for heat processing equipment can utilize such data to adjust gas input modes to provide a more uniform total heat input. They can also utilize such information to adjust burner controls so as to provide the proper air-fuel ratio and thereby avoid inefficiencies which occur when the burners are operating at a ratio which is either too lean or too rich.
Attempts have been made to provide "on-line" devices. In the prior art it is generally known that the heat content of a gaseous fuel such as natural gas is related to the ratio of fuel to air necessary for complete combustion of the gas. One prior art device utilizes a system in which the heat content of the fuel is related mathematically to that ratio of air to fuel which maximizes the adiabatic flame temperature of the mixture. In that system the air and fuel are split between two burners in such a way that the mixture in one burner has a slightly higher air-fuel ratio than the other. The air flow is allowed to remain constant, and the fuel flow is varied in the two burners until the temperature of both flames is the same. In this manner, one burner approaches the maximum flame temperature from the "rich" side and the other from the "lean" side. The fuel flow is metered as by a turbine meter, and, based on the metered rates of air and fuel, a BTU heat content is calculated assuming the maximum adiabatic flame temperature occurs at the stoichiometric point of the combustion mixture. Actually, the maximum flame temperature may be reached when the mixture is somewhat lean, i.e. when an amount of excess air is present.
A prior art method of calculating the heat content from such measurements is found in "New Approach to the Continuous Measurement of Calorific Values of Gaseous Fuels" by William H. Clingman, Jr., AGA Operating Section (1972). That system has the drawback that it requires both accurate fuel measurement and temperature measurement of two complete burner systems for comparison.
The use of electrochemical cells as oxygen or combustibles sensors to sense the residual products of combustion including electrochemical cells based on ceramic compounds such as ZrO.sub.2 is also known. Under normal conditions, natural gas is burned with an excess of oxygen to assure complete combustion and an absence of carbon monoxide in the products of combustion. This leads to the presence of an amount of excess oxygen after combustion which raises the possibility that an oxygen sensor based on zirconia or the like could provide a relatively inexpensive and rapid solution to the problem of determining heating values for natural gas mixtures based on the air-fuel ratio. Such solid electrolyte-based oxygen sensors have been used to rapidly provide information for such systems as catalytic automobile exhaust control. Several different types of solid electrolyte-based oxygen sensors are commercially available.
Thus it has been proposed to use a ceramic based electrochemical made chiefly of zirconia (ZrO.sub.2) which is known to exhibit a Nernstian voltage output when exposed to differing partial pressures of oxygen on each side of the ceramic material. This can be used to sense the amount of oxygen present in the products of combustion. That system proposes to utilize a known mass ratio between the fuel and oxygen supplied to a burner in conjunction with the measurement of excess oxygen after combustion using the Nernstian relationship to provide a basis for deriving the heat content of the fuel. According to that system, measurements should be made when the combustion mixture contains about 20 percent excess air, that is, not stoichiometric combustion.
That system requires that the air and fuel be controlled so as to be measured in relation to standard temperature and pressure and precise metering of the mass flows of fuel and air. Also the sensor output is decidedly temperature sensitive in the presence of more than minute amounts of oxygen. Thus, when operating in the oxygen-rich portion of the ZrO.sub.2 electrochemical response curve, the temperature of the sensor must be carefully controlled. Unfortunately, the system does not take into consideration the fact that as the heating value of the gas to be measured increases, so does the average molecular weight of the gas. While this could be overcome with appropriate correction factors, if only combustible gases were involved, the molecular weight of the fuel species can also be varied because of the presence of inert species in the mixture which leads to calculation errors.